marofmar commited on
Commit ·
2f3abc7
1
Parent(s): 2ee585c
checkpoint 84 model and tokenizer
Browse files- config.json +76 -0
- optimizer.pt +3 -0
- preprocessor_config.json +9 -0
- pytorch_model.bin +3 -0
- scheduler.pt +3 -0
- trainer_state.json +310 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,76 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "facebook/wav2vec2-large-xlsr-53",
|
| 3 |
+
"activation_dropout": 0.0,
|
| 4 |
+
"apply_spec_augment": true,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"Wav2Vec2ForCTC"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.1,
|
| 9 |
+
"bos_token_id": 1,
|
| 10 |
+
"conv_bias": true,
|
| 11 |
+
"conv_dim": [
|
| 12 |
+
512,
|
| 13 |
+
512,
|
| 14 |
+
512,
|
| 15 |
+
512,
|
| 16 |
+
512,
|
| 17 |
+
512,
|
| 18 |
+
512
|
| 19 |
+
],
|
| 20 |
+
"conv_kernel": [
|
| 21 |
+
10,
|
| 22 |
+
3,
|
| 23 |
+
3,
|
| 24 |
+
3,
|
| 25 |
+
3,
|
| 26 |
+
2,
|
| 27 |
+
2
|
| 28 |
+
],
|
| 29 |
+
"conv_stride": [
|
| 30 |
+
5,
|
| 31 |
+
2,
|
| 32 |
+
2,
|
| 33 |
+
2,
|
| 34 |
+
2,
|
| 35 |
+
2,
|
| 36 |
+
2
|
| 37 |
+
],
|
| 38 |
+
"ctc_loss_reduction": "mean",
|
| 39 |
+
"ctc_zero_infinity": false,
|
| 40 |
+
"do_stable_layer_norm": true,
|
| 41 |
+
"eos_token_id": 2,
|
| 42 |
+
"feat_extract_activation": "gelu",
|
| 43 |
+
"feat_extract_dropout": 0.0,
|
| 44 |
+
"feat_extract_norm": "layer",
|
| 45 |
+
"feat_proj_dropout": 0.0,
|
| 46 |
+
"final_dropout": 0.0,
|
| 47 |
+
"gradient_checkpointing": true,
|
| 48 |
+
"hidden_act": "gelu",
|
| 49 |
+
"hidden_dropout": 0.1,
|
| 50 |
+
"hidden_size": 1024,
|
| 51 |
+
"initializer_range": 0.02,
|
| 52 |
+
"intermediate_size": 4096,
|
| 53 |
+
"layer_norm_eps": 1e-05,
|
| 54 |
+
"layerdrop": 0.1,
|
| 55 |
+
"mask_channel_length": 10,
|
| 56 |
+
"mask_channel_min_space": 1,
|
| 57 |
+
"mask_channel_other": 0.0,
|
| 58 |
+
"mask_channel_prob": 0.0,
|
| 59 |
+
"mask_channel_selection": "static",
|
| 60 |
+
"mask_feature_length": 10,
|
| 61 |
+
"mask_feature_prob": 0.0,
|
| 62 |
+
"mask_time_length": 10,
|
| 63 |
+
"mask_time_min_space": 1,
|
| 64 |
+
"mask_time_other": 0.0,
|
| 65 |
+
"mask_time_prob": 0.05,
|
| 66 |
+
"mask_time_selection": "static",
|
| 67 |
+
"model_type": "wav2vec2",
|
| 68 |
+
"num_attention_heads": 16,
|
| 69 |
+
"num_conv_pos_embedding_groups": 16,
|
| 70 |
+
"num_conv_pos_embeddings": 128,
|
| 71 |
+
"num_feat_extract_layers": 7,
|
| 72 |
+
"num_hidden_layers": 24,
|
| 73 |
+
"pad_token_id": 259,
|
| 74 |
+
"transformers_version": "4.6.0.dev0",
|
| 75 |
+
"vocab_size": 260
|
| 76 |
+
}
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5bfc6508f4174c9f6359349b9687c28ddab0e09aa92126ab0fa9bcf89e6bb4fb
|
| 3 |
+
size 2492208775
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_normalize": true,
|
| 3 |
+
"feature_extractor_type": "Wav2Vec2FeatureExtractor",
|
| 4 |
+
"feature_size": 1,
|
| 5 |
+
"padding_side": "right",
|
| 6 |
+
"padding_value": 0.0,
|
| 7 |
+
"return_attention_mask": true,
|
| 8 |
+
"sampling_rate": 16000
|
| 9 |
+
}
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:81a8acfe830487f7ddb00a2c4c6db079bdc50bf9d4ba7f6fd18f42934341b452
|
| 3 |
+
size 1262999831
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:057999ec38e996dec146de5e28f3aaf30dfbc86f0c54eee12d611f5e3b7423e8
|
| 3 |
+
size 623
|
trainer_state.json
ADDED
|
@@ -0,0 +1,310 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 4.255084991455078,
|
| 3 |
+
"best_model_checkpoint": "./xlsr-demo/checkpoint-80",
|
| 4 |
+
"epoch": 28.0,
|
| 5 |
+
"global_step": 84,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 1.33,
|
| 12 |
+
"learning_rate": 5.9999999999999995e-05,
|
| 13 |
+
"loss": 45.1762,
|
| 14 |
+
"step": 4
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 1.33,
|
| 18 |
+
"eval_loss": 32.95661544799805,
|
| 19 |
+
"eval_runtime": 0.6944,
|
| 20 |
+
"eval_samples_per_second": 14.402,
|
| 21 |
+
"eval_wer": 1.0,
|
| 22 |
+
"step": 4
|
| 23 |
+
},
|
| 24 |
+
{
|
| 25 |
+
"epoch": 2.67,
|
| 26 |
+
"learning_rate": 0.00023999999999999998,
|
| 27 |
+
"loss": 48.4094,
|
| 28 |
+
"step": 8
|
| 29 |
+
},
|
| 30 |
+
{
|
| 31 |
+
"epoch": 2.67,
|
| 32 |
+
"eval_loss": 32.567665100097656,
|
| 33 |
+
"eval_runtime": 0.5168,
|
| 34 |
+
"eval_samples_per_second": 19.351,
|
| 35 |
+
"eval_wer": 1.0,
|
| 36 |
+
"step": 8
|
| 37 |
+
},
|
| 38 |
+
{
|
| 39 |
+
"epoch": 4.0,
|
| 40 |
+
"learning_rate": 0.00028941176470588233,
|
| 41 |
+
"loss": 41.452,
|
| 42 |
+
"step": 12
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"epoch": 4.0,
|
| 46 |
+
"eval_loss": 31.335033416748047,
|
| 47 |
+
"eval_runtime": 0.5187,
|
| 48 |
+
"eval_samples_per_second": 19.279,
|
| 49 |
+
"eval_wer": 1.0,
|
| 50 |
+
"step": 12
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 5.33,
|
| 54 |
+
"learning_rate": 0.0002788235294117647,
|
| 55 |
+
"loss": 43.7158,
|
| 56 |
+
"step": 16
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 5.33,
|
| 60 |
+
"eval_loss": 29.559152603149414,
|
| 61 |
+
"eval_runtime": 0.5095,
|
| 62 |
+
"eval_samples_per_second": 19.627,
|
| 63 |
+
"eval_wer": 1.0,
|
| 64 |
+
"step": 16
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 6.67,
|
| 68 |
+
"learning_rate": 0.00026470588235294115,
|
| 69 |
+
"loss": 30.5266,
|
| 70 |
+
"step": 20
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 6.67,
|
| 74 |
+
"eval_loss": 18.15114974975586,
|
| 75 |
+
"eval_runtime": 0.5292,
|
| 76 |
+
"eval_samples_per_second": 18.896,
|
| 77 |
+
"eval_wer": 1.0,
|
| 78 |
+
"step": 20
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 8.0,
|
| 82 |
+
"learning_rate": 0.0002505882352941176,
|
| 83 |
+
"loss": 27.451,
|
| 84 |
+
"step": 24
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 8.0,
|
| 88 |
+
"eval_loss": 13.7584810256958,
|
| 89 |
+
"eval_runtime": 0.5073,
|
| 90 |
+
"eval_samples_per_second": 19.713,
|
| 91 |
+
"eval_wer": 1.0,
|
| 92 |
+
"step": 24
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 9.33,
|
| 96 |
+
"learning_rate": 0.0002364705882352941,
|
| 97 |
+
"loss": 19.4205,
|
| 98 |
+
"step": 28
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 9.33,
|
| 102 |
+
"eval_loss": 10.164986610412598,
|
| 103 |
+
"eval_runtime": 0.5127,
|
| 104 |
+
"eval_samples_per_second": 19.503,
|
| 105 |
+
"eval_wer": 1.0,
|
| 106 |
+
"step": 28
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 10.67,
|
| 110 |
+
"learning_rate": 0.00022235294117647057,
|
| 111 |
+
"loss": 13.2132,
|
| 112 |
+
"step": 32
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 10.67,
|
| 116 |
+
"eval_loss": 7.898101806640625,
|
| 117 |
+
"eval_runtime": 0.5031,
|
| 118 |
+
"eval_samples_per_second": 19.877,
|
| 119 |
+
"eval_wer": 1.0,
|
| 120 |
+
"step": 32
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 12.0,
|
| 124 |
+
"learning_rate": 0.00020823529411764704,
|
| 125 |
+
"loss": 9.3548,
|
| 126 |
+
"step": 36
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 12.0,
|
| 130 |
+
"eval_loss": 6.538975715637207,
|
| 131 |
+
"eval_runtime": 0.5312,
|
| 132 |
+
"eval_samples_per_second": 18.825,
|
| 133 |
+
"eval_wer": 1.0,
|
| 134 |
+
"step": 36
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 13.33,
|
| 138 |
+
"learning_rate": 0.0001941176470588235,
|
| 139 |
+
"loss": 8.7253,
|
| 140 |
+
"step": 40
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 13.33,
|
| 144 |
+
"eval_loss": 5.714229583740234,
|
| 145 |
+
"eval_runtime": 0.4983,
|
| 146 |
+
"eval_samples_per_second": 20.07,
|
| 147 |
+
"eval_wer": 1.0,
|
| 148 |
+
"step": 40
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 14.67,
|
| 152 |
+
"learning_rate": 0.00017999999999999998,
|
| 153 |
+
"loss": 6.8295,
|
| 154 |
+
"step": 44
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 14.67,
|
| 158 |
+
"eval_loss": 5.214608192443848,
|
| 159 |
+
"eval_runtime": 0.5112,
|
| 160 |
+
"eval_samples_per_second": 19.561,
|
| 161 |
+
"eval_wer": 1.0,
|
| 162 |
+
"step": 44
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 16.0,
|
| 166 |
+
"learning_rate": 0.00016588235294117646,
|
| 167 |
+
"loss": 5.9458,
|
| 168 |
+
"step": 48
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 16.0,
|
| 172 |
+
"eval_loss": 4.887683868408203,
|
| 173 |
+
"eval_runtime": 0.5156,
|
| 174 |
+
"eval_samples_per_second": 19.394,
|
| 175 |
+
"eval_wer": 1.0,
|
| 176 |
+
"step": 48
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 17.33,
|
| 180 |
+
"learning_rate": 0.00015176470588235293,
|
| 181 |
+
"loss": 5.4214,
|
| 182 |
+
"step": 52
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 17.33,
|
| 186 |
+
"eval_loss": 4.679065704345703,
|
| 187 |
+
"eval_runtime": 1.0634,
|
| 188 |
+
"eval_samples_per_second": 9.403,
|
| 189 |
+
"eval_wer": 1.0,
|
| 190 |
+
"step": 52
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 18.67,
|
| 194 |
+
"learning_rate": 0.0001376470588235294,
|
| 195 |
+
"loss": 5.8716,
|
| 196 |
+
"step": 56
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 18.67,
|
| 200 |
+
"eval_loss": 4.548101425170898,
|
| 201 |
+
"eval_runtime": 0.4987,
|
| 202 |
+
"eval_samples_per_second": 20.051,
|
| 203 |
+
"eval_wer": 1.0,
|
| 204 |
+
"step": 56
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 20.0,
|
| 208 |
+
"learning_rate": 0.00012352941176470587,
|
| 209 |
+
"loss": 4.69,
|
| 210 |
+
"step": 60
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 20.0,
|
| 214 |
+
"eval_loss": 4.435084342956543,
|
| 215 |
+
"eval_runtime": 0.6371,
|
| 216 |
+
"eval_samples_per_second": 15.695,
|
| 217 |
+
"eval_wer": 1.0,
|
| 218 |
+
"step": 60
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 21.33,
|
| 222 |
+
"learning_rate": 0.00010941176470588234,
|
| 223 |
+
"loss": 4.6092,
|
| 224 |
+
"step": 64
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 21.33,
|
| 228 |
+
"eval_loss": 4.350079536437988,
|
| 229 |
+
"eval_runtime": 0.5165,
|
| 230 |
+
"eval_samples_per_second": 19.361,
|
| 231 |
+
"eval_wer": 1.0,
|
| 232 |
+
"step": 64
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 22.67,
|
| 236 |
+
"learning_rate": 9.529411764705882e-05,
|
| 237 |
+
"loss": 4.9462,
|
| 238 |
+
"step": 68
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 22.67,
|
| 242 |
+
"eval_loss": 4.301529884338379,
|
| 243 |
+
"eval_runtime": 0.4931,
|
| 244 |
+
"eval_samples_per_second": 20.278,
|
| 245 |
+
"eval_wer": 1.0,
|
| 246 |
+
"step": 68
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 24.0,
|
| 250 |
+
"learning_rate": 8.117647058823529e-05,
|
| 251 |
+
"loss": 4.7406,
|
| 252 |
+
"step": 72
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 24.0,
|
| 256 |
+
"eval_loss": 4.267996311187744,
|
| 257 |
+
"eval_runtime": 0.5187,
|
| 258 |
+
"eval_samples_per_second": 19.279,
|
| 259 |
+
"eval_wer": 1.0,
|
| 260 |
+
"step": 72
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 25.33,
|
| 264 |
+
"learning_rate": 6.705882352941176e-05,
|
| 265 |
+
"loss": 4.5484,
|
| 266 |
+
"step": 76
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 25.33,
|
| 270 |
+
"eval_loss": 4.258701801300049,
|
| 271 |
+
"eval_runtime": 0.5059,
|
| 272 |
+
"eval_samples_per_second": 19.768,
|
| 273 |
+
"eval_wer": 1.0,
|
| 274 |
+
"step": 76
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 26.67,
|
| 278 |
+
"learning_rate": 5.294117647058824e-05,
|
| 279 |
+
"loss": 4.2841,
|
| 280 |
+
"step": 80
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 26.67,
|
| 284 |
+
"eval_loss": 4.255084991455078,
|
| 285 |
+
"eval_runtime": 0.493,
|
| 286 |
+
"eval_samples_per_second": 20.286,
|
| 287 |
+
"eval_wer": 1.0,
|
| 288 |
+
"step": 80
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 28.0,
|
| 292 |
+
"learning_rate": 3.882352941176471e-05,
|
| 293 |
+
"loss": 4.4733,
|
| 294 |
+
"step": 84
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 28.0,
|
| 298 |
+
"eval_loss": 4.255660533905029,
|
| 299 |
+
"eval_runtime": 0.5041,
|
| 300 |
+
"eval_samples_per_second": 19.839,
|
| 301 |
+
"eval_wer": 1.0,
|
| 302 |
+
"step": 84
|
| 303 |
+
}
|
| 304 |
+
],
|
| 305 |
+
"max_steps": 90,
|
| 306 |
+
"num_train_epochs": 30,
|
| 307 |
+
"total_flos": 3.564185651712e+17,
|
| 308 |
+
"trial_name": null,
|
| 309 |
+
"trial_params": null
|
| 310 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2c2220a89d8cda2c9e9ddfe7fd922e39268d2ceebcd96397807b4516c8d2e156
|
| 3 |
+
size 2351
|